2004
@inproceedings{Fay2004,
vgclass = {refpap},
author = {Usama Fayyad},
title = {Data Mining Grand Challenges},
booktitle = {Proceedings of the 8th Pacific-Asia Conference on Advances
in Knowledge Discovery and Data Mining (PAKDD 2004)},
address = {Sydney, Australia},
number = {3056},
series = {Lecture Notes in Computer Science},
pages = {2},
publisher = {Springer-Verlag},
month = {May~26--28},
year = {2004},
note = {(keynote speech)},
url = {http://springerlink.metapress.com.ezproxy.lib.monash.edu.au/link.asp?id=qc5crk1jdljff9qq},
abstract = {The past two decades has seen a huge wave of computational
systems for the ``digitization'' of business operations from ERP,
to manufacturing, to systems for customer interactions. These systems
increased the throughput and efficiency of conducting
ldquotransactionsrdquo and resulted in an unprecedented build-up of
data captured from these systems. The paradoxical reality that most
organizations face today is that they have more data about every aspect
of their operations and customers, yet they find themselves with an
ever diminishing understanding of either. Data Mining has received much
attention as a technology that can possibly bridge the gap between data
and knowledge.
While some interesting progress has been achieved over the past few
years, especially when it comes to techniques and scalable algorithms,
very few organizations have managed to bene t from the technology.
Despite the recent advances, some major hurdles exist on the road to
the needed evolution. Furthermore, most technical research work does
not appear to be directed at these challenges, nor does it appear to be
aware of their nature. This talk will cover these challenges and
present them in both the technical and the business context. The
exposition will cover deep technical research questions, practical
application considerations, and social/economic considerations. The
talk will draw on illustrative examples from scienti c data analysis,
commercial applications of data mining in understanding customer
interaction data, and considerations of coupling data mining technology
within database management of systems. Of particular interest is the
business challenge of how to make the technology really work in
practice. There are many unsolved deep technical research problems in
this eld and we conclude by covering a sampling of these.},
}